+/- for the Misfire against Michigan State

[PurpleWhiteBoy]

Not a lot of happy thoughts from the numbers today... Luke Hunger didn't play and a big Spartan run put NU in a 47-28 hole at halftime.
I scrapped the last 46 seconds. We were down 69-54 with 4 minutes to play, which is close to garbage time, but not quite.
The lineup of Fitzmorris/Martinelli/Barnhizer/Leach/Windham played from 4:18 until 46 seconds, outscoring the Spartans 8-2, helping each of those 5 guys in the table below.

Player	Minutes	NU Pts	MSU Pts	Raw +'-	Player +/-	Net +/-	Box Pts
Martinelli	37	62	69	-7	+5.29	+3.89	9.65
Nicholson	21	31	34	-3	+2.36	+1.76	4.80
Leach	34	59	58	+1	-0.04	+0.16	6.30
Windham	9.5	19	17	+2	-0.84	-0.44	0.60
Ciaravino	7.5	8	12	-4	-0.14	-0.94	2.00
Fitzmorris	18	31	35	-4	-1.29	-2.09	0.45
Mullins	16	27	34	-7	-1.13	-2.53	1.30
Berry	16	17	29	-12	-0.95	-3.35	-0.10
Barnhizer	37	56	67	-11	-3.26	-5.46	0.40

Nick Martinelli played a good game with a little help from Jalen Leach and Matt Nicholson. Nobody else did too much. Brooks Barnhizer played his 2nd bad game in a row ,going 2 of 13 from the field, including 0 of 6 from distance. Ty Berry struggled again.

The starters played to a 12-12 draw in 9 minutes of court time. Defense, not much scoring.
Lineup #2 (now Fitzmorris/Martinelli/Barnhizer/Mullins/Leach) played to a 16-15 deficit in 6:06. Scoring, not much defense.

When Berry was out there with Leach we got outscored 18-17 in about 13 minutes.
When Berry was out there with Mullins, we got smoked 9-0 in a mere 3 minutes. Mullins turned it over twice in a 7-0 run for the Spartans with Fitzmorris, Martinelli and Barnhizer also on the court.

Ultimately, NU's fortunes have changed dramatically in conference play. The starting 5 has gone from highly effective to ineffective. I believe that reflects the fact that Big Ten coaches are very familiar with our players and our approach to the game. Non-conference opponents played us more or less "straight up." Big Ten coaches are smothering Barnhizer and leaving Mullins and Hunger alone on the perimeter, ignoring Nicholson. Our starters were much more effective on the offensive end when we held a physical advantage over our opponents. That will rarely be the case in conference play.

I'd like to see Ciaravino or Windham out there for Berry as our first substitution, with Nicholson staying in the game. Ty is 2 for 16 (total) in our last 4 Big Ten games. Unfortunately thats more of a "hope" than an expectation.

 

[EagerFan]

Windham and Ciaravino both hit a three from the corner last game (1-3 and1-2 iirc) and Angelo had that nice put back dunk at the end. I'd like to think they'll both get a little more time when the starters are shanking threes.

 

[PurpleWhiteBoy]

Windham and Ciaravino both hit a three from the corner last game (1-3 and1-2 iirc) and Angelo had that nice put back dunk at the end. I'd like to think they'll both get a little more time when the starters are shanking threes.

They both impress me.
Ciaravino is explosive coming down the baseline for offensive rebounds and/or dunks. High impact on offense. Multi-level scorer. Pretty good rebounder.

It seems fairly clear that Barnhizer is struggling because he has to run the offense. It is also pretty clear that Leach (and of course Berry) can't run the offense much of the time. So that screams for Windham. He is the only guy on the team who can dribble the ball while facing the opposing point guard - Barnhizer frequently has to turn his shoulder to prevent getting his pocket picked. Makes it hard to see the court.

Having said that, it is almost mandatory that Collins try Nicholson/Martinelli/Barnhizer/Leach/Ciaravino.
Barnhizer runs the offense with 3 legit scoring options around him.

 

[Deeringfish]

Honestly I think Berry should have taken a medical shirt if that would be allowed. He obviously hasn't recovered from his injury well enough to perform at this level. I think Collins should try making him the 6th man. Maybe that would help him be a little fresher and give his legs a little boost.

 

[EagerFan]

...Windham. He is the only guy on the team who can dribble the ball while facing the opposing point guard - Barnhizer frequently has to turn his shoulder to prevent getting his pocket picked. Makes it hard to see the court.

This is definitely a problem for the offense.

There are some guys Leach can go face up against, but not too many in conference. He is a solid shooting guard and operating as point is not his best usage.

PSU used that against Brooks a bunch. Their two guard waiting until Brooks got a little past half court and line of sight on him, then circling around behind to try and steal as Brooks was turned away from center court to protect the ball from his defender. Had some success, unfortunately.

 

[prez77]

Not a lot of happy thoughts from the numbers today... Luke Hunger didn't play and a big Spartan run put NU in a 47-28 hole at halftime.
I scrapped the last 46 seconds. We were down 69-54 with 4 minutes to play, which is close to garbage time, but not quite.
The lineup of Fitzmorris/Martinelli/Barnhizer/Leach/Windham played from 4:18 until 46 seconds, outscoring the Spartans 8-2, helping each of those 5 guys in the table below.

Player	Minutes	NU Pts	MSU Pts	Raw +'-	Player +/-	Net +/-	Box Pts
Martinelli	37	62	69	-7	+5.29	+3.89	9.65
Nicholson	21	31	34	-3	+2.36	+1.76	4.80
Leach	34	59	58	+1	-0.04	+0.16	6.30
Windham	9.5	19	17	+2	-0.84	-0.44	0.60
Ciaravino	7.5	8	12	-4	-0.14	-0.94	2.00
Fitzmorris	18	31	35	-4	-1.29	-2.09	0.45
Mullins	16	27	34	-7	-1.13	-2.53	1.30
Berry	16	17	29	-12	-0.95	-3.35	-0.10
Barnhizer	37	56	67	-11	-3.26	-5.46	0.40

Nick Martinelli played a good game with a little help from Jalen Leach and Matt Nicholson. Nobody else did too much. Brooks Barnhizer played his 2nd bad game in a row ,going 2 of 13 from the field, including 0 of 6 from distance. Ty Berry struggled again.

The starters played to a 12-12 draw in 9 minutes of court time. Defense, not much scoring.
Lineup #2 (now Fitzmorris/Martinelli/Barnhizer/Mullins/Leach) played to a 16-15 deficit in 6:06. Scoring, not much defense.

When Berry was out there with Leach we got outscored 18-17 in about 13 minutes.
When Berry was out there with Mullins, we got smoked 9-0 in a mere 3 minutes. Mullins turned it over twice in a 7-0 run for the Spartans with Fitzmorris, Martinelli and Barnhizer also on the court.

Ultimately, NU's fortunes have changed dramatically in conference play. The starting 5 has gone from highly effective to ineffective. I believe that reflects the fact that Big Ten coaches are very familiar with our players and our approach to the game. Non-conference opponents played us more or less "straight up." Big Ten coaches are smothering Barnhizer and leaving Mullins and Hunger alone on the perimeter, ignoring Nicholson. Our starters were much more effective on the offensive end when we held a physical advantage over our opponents. That will rarely be the case in conference play.

I'd like to see Ciaravino or Windham out there for Berry as our first substitution, with Nicholson staying in the game. Ty is 2 for 16 (total) in our last 4 Big Ten games. Unfortunately thats more of a "hope" than an expectation.

Have much appreciated your +/- analysis. Great stuff. Want to run an idea by you and any other statisticians on the Board.

If I were approaching a problem like this in some areas I have worked in professionally, I would take a data set that has a record for each NU possession and have a variable for the type of shot that was made, a 3, 2, 1 or 0 and a dummy variable for each player who played in the game, NU and the opponent. I would then regress the player variable, essentially on the floor or not, with the points outcome of the possession. The resulting coefficients for each of the dummies would indicate their relative contributions to scoring process. This method would have the virtue of controlling for the player composition of our opponent's defense on each possession as well - shooting the ball well matters, but who is trying to block the shot also matters. I think this calculation would address offense.

For defense, you could either reverse the process for the opponent's possessions, or set up the data base for all game possessions and add -3, -2, or -1 for the outcome variables to represent the opponent's scores. This should result in a coefficient for each player describing their combined contribution to NU scoring and their association with however much the opponent scores for a net player value.

I can imagine some problems getting meaningful estimates for guys who play 40 minutes or for 1 or 2 minutes. If there were enough of the 40 minute guys, it might blow up the equation as can happen in a regression where the dependent values don't align at all with one or more of the independent variables. But there might be some work-arounds. Interested in people's thoughts/critique.

 

[macarthur31]

Have much appreciated your +/- analysis. Great stuff. Want to run an idea by you and any other statisticians on the Board.

If I were approaching a problem like this in some areas I have worked in professionally, I would take a data set that has a record for each NU possession and have a variable for the type of shot that was made, a 3, 2, 1 or 0 and a dummy variable for each player who played in the game, NU and the opponent. I would then regress the player variable, essentially on the floor or not, with the points outcome of the possession. The resulting coefficients for each of the dummies would indicate their relative contributions to scoring process. This method would have the virtue of controlling for the player composition of our opponent's defense on each possession as well - shooting the ball well matters, but who is trying to block the shot also matters. I think this calculation would address offense.

For defense, you could either reverse the process for the opponent's possessions, or set up the data base for all game possessions and add -3, -2, or -1 for the outcome variables to represent the opponent's scores. This should result in a coefficient for each player describing their combined contribution to NU scoring and their association with however much the opponent scores for a net player value.

I can imagine some problems getting meaningful estimates for guys who play 40 minutes or for 1 or 2 minutes. If there were enough of the 40 minute guys, it might blow up the equation as can happen in a regression where the dependent values don't align at all with one or more of the independent variables. But there might be some work-arounds. Interested in people's thoughts/critique.

Whoa. I'm just a dude with a kenpom subsscription. Reading that humbled me and reminded me that I was Comm Studies back in the day. I'll defer to other hard sciences folks on this method.

 

[Purple Pile Driver]

Have much appreciated your +/- analysis. Great stuff. Want to run an idea by you and any other statisticians on the Board.

If I were approaching a problem like this in some areas I have worked in professionally, I would take a data set that has a record for each NU possession and have a variable for the type of shot that was made, a 3, 2, 1 or 0 and a dummy variable for each player who played in the game, NU and the opponent. I would then regress the player variable, essentially on the floor or not, with the points outcome of the possession. The resulting coefficients for each of the dummies would indicate their relative contributions to scoring process. This method would have the virtue of controlling for the player composition of our opponent's defense on each possession as well - shooting the ball well matters, but who is trying to block the shot also matters. I think this calculation would address offense.

For defense, you could either reverse the process for the opponent's possessions, or set up the data base for all game possessions and add -3, -2, or -1 for the outcome variables to represent the opponent's scores. This should result in a coefficient for each player describing their combined contribution to NU scoring and their association with however much the opponent scores for a net player value.

I can imagine some problems getting meaningful estimates for guys who play 40 minutes or for 1 or 2 minutes. If there were enough of the 40 minute guys, it might blow up the equation as can happen in a regression where the dependent values don't align at all with one or more of the independent variables. But there might be some work-arounds. Interested in people's thoughts/critique.

I have no idea what you said, but it sounds great to me!

 

[prez77]

I have no idea what you said, but it sounds great to me!

Won't belabor, but let's say we wanted to understand what drives economic growth instead of scoring growth. We would set up a data base that had the dependent variable of annual growth over some period of years (rather than scoring by possession) and match those years with possible reasons for that growth like each year's energy prices, steel prices, college graduates, tax rates, new patents etc. (rather than which players happen to be on the floor). We would then use a statistical technique on our data base called a regression that assesses the relative contribution of the various economic inputs on the output, growth. So, as I was pondering macarthur31's data, I found myself wondering if the same econometric methods would apply to explaining scoring output.

Scoring is the result of a. bunch of inputs, the shooter, his helpers (point guards, screeners etc.) and defenders who fail to defend. They all contribute to varying degrees to every point scored. Who is playing defense really, really matters. There's a profound difference between charging into the lane and confronting the last guy on the bench as opposed to Wemby or Big Matt. It's very measurable.

 

[PurpleWhiteBoy]

Have much appreciated your +/- analysis. Great stuff. Want to run an idea by you and any other statisticians on the Board.

If I were approaching a problem like this in some areas I have worked in professionally, I would take a data set that has a record for each NU possession and have a variable for the type of shot that was made, a 3, 2, 1 or 0 and a dummy variable for each player who played in the game, NU and the opponent. I would then regress the player variable, essentially on the floor or not, with the points outcome of the possession. The resulting coefficients for each of the dummies would indicate their relative contributions to scoring process. This method would have the virtue of controlling for the player composition of our opponent's defense on each possession as well - shooting the ball well matters, but who is trying to block the shot also matters. I think this calculation would address offense.

For defense, you could either reverse the process for the opponent's possessions, or set up the data base for all game possessions and add -3, -2, or -1 for the outcome variables to represent the opponent's scores. This should result in a coefficient for each player describing their combined contribution to NU scoring and their association with however much the opponent scores for a net player value.

I can imagine some problems getting meaningful estimates for guys who play 40 minutes or for 1 or 2 minutes. If there were enough of the 40 minute guys, it might blow up the equation as can happen in a regression where the dependent values don't align at all with one or more of the independent variables. But there might be some work-arounds. Interested in people's thoughts/critique.

I'm glad you like my +/- breakdowns and appreciate the interest!

It seems like you're proposing an improvement on what I call "Raw +/-"
The first thing I should mention is that if you are going to look at the capabilities of defenders on the other team, your data requirements just exploded by a factor of 1000 or so - because you have to do every player for every team for each of their games! I don't look at anybody but the players on our team. There are stats/programming guys like Evan Miyakawa who (I believe) are doing the whole universe - breaking every game down based on a possession, a result and what 10 players are on the court. Its very ambitious, but certainly impressive. I don't think Torvik and KenPom are doing their analytics based on which 5 guys were on the court at any given time (though I could be wrong). My impression is that they deal in total points and total possessions, which is much, much simpler.

If you look only at the results of a possession (4,3,2,1,0 points) and who was on the court for us, it seems to me that you are implicitly assuming that everybody shares equally in success or failure AND that any combination of 5 guys is just as likely to succeed as another.
So if Nicholson, Hunger, Fitzmorris, Barnhizer and Martinelli were deployed together and we got outscored 15-0, that would reflect poorly on each of them, when really it was just a crazy move by the coach.

So you have to try to assign credit or blame. Who missed the shots? Who turned the ball over? Who hit the 3 pointer? (Thats what I try to do with my Player +/-)

Admittedly what you are proposing would be an improvement - I'd probably look at defense as a 5 man unit instead of 5 individuals - but there is a lot of noise in the data, so all that extra complexity might not get you much.

 

[prez77]

I'm glad you like my +/- breakdowns and appreciate the interest!

It seems like you're proposing an improvement on what I call "Raw +/-"
The first thing I should mention is that if you are going to look at the capabilities of defenders on the other team, your data requirements just exploded by a factor of 1000 or so - because you have to do every player for every team for each of their games! I don't look at anybody but the players on our team. There are stats/programming guys like Evan Miyakawa who (I believe) are doing the whole universe - breaking every game down based on a possession, a result and what 10 players are on the court. Its very ambitious, but certainly impressive. I don't think Torvik and KenPom are doing their analytics based on which 5 guys were on the court at any given time (though I could be wrong). My impression is that they deal in total points and total possessions, which is much, much simpler.

If you look only at the results of a possession (4,3,2,1,0 points) and who was on the court for us, it seems to me that you are implicitly assuming that everybody shares equally in success or failure AND that any combination of 5 guys is just as likely to succeed as another.
So if Nicholson, Hunger, Fitzmorris, Barnhizer and Martinelli were deployed together and we got outscored 15-0, that would reflect poorly on each of them, when really it was just a crazy move by the coach.

So you have to try to assign credit or blame. Who missed the shots? Who turned the ball over? Who hit the 3 pointer? (Thats what I try to do with my Player +/-)

Admittedly what you are proposing would be an improvement - I'd probably look at defense as a 5 man unit instead of 5 individuals - but there is a lot of noise in the data, so all that extra complexity might not get you much.

Sort of. Using a regression doesn't value anything except what is out there when the ball goes through the hoop, or doesn't. It "assumes" as I somewhat do, that scoring usually requires all 5 guys, and that 5 guys participate to varying degrees in a failure to defend against a made basket. So under the method I explore, it requires only the data for a particular game and, like you, just asks who seems to be on the floor when shots are made or not made and produces numbers for each player based on those combinations. It would tend to replicate the Raw +/-, but would also account for defender +/- in the game.

The crazy coach lineup you suggested is, of course, crazy because it's not a useful combination of skills - which is what your analysis is all about - the best-producing groups. In a way I'm defending the Raw +/- because yes, Martinelli puts the ball in the basket but he does it more often if someone screens for him and the defense has to account for one or two guys who will make a 3 if they pay too much attention to Martinelli. Most baskets, whoever shoots them, are team accomplishments once you're in the half-court game unless you've got Curry from 30 feet, or Jordan or James just overpowering the opposition. And by the same token, poor defense can make someone who doesn't really shoot that well look a lot better. That's why our offense is so much better against the Tier 3s and 4s we play early in the season.